A Salt and Pepper Noise Image Filtering Method Using PCNN

Rencan Nie,Shaowen Yao,Dongming Zhou,Xiang Li
DOI: https://doi.org/10.1007/978-3-642-37502-6_120
2013-01-01
Abstract:Based on Pulse Coupled Neural networks, an effective salt and pepper noise image filtering method is proposed. Similar groups of neurons burst synchronous pulses in a PCNN, thereby the noise pixels are detected. Considering a noise pixel has the most similarity with neighbor non-noise pixels, a filtering method called extended median filtering is put forward to filter the noises in an image. Simulation results show that the method proposed has excellent filtering performance for the noise images with different noise intensity, and has the obvious advantage compared with the median filter.
What problem does this paper attempt to address?